Michael Hsueh and Grant Sporre set out the landscape for oil demand and supply over the course of 2017/18 as all eyes remain fixed on OPEC and Saudi Arabia
After two years of supranormal growth in a declining oil price environment (down by 46% in 2015 and 16% in 2016) we expect demand growth for oil closer to the PPP-weighted (purchasing power parity) world GDP model as oil prices rise by 22% by our forecast.
As a result of an upgrade to European demand growth from a decline of 70,000 barrels per day (b/d) to an identical growth of 70,000 b/d, we now expect total demand growth of 1,385 million b/d in 2017, equal to GDP model growth of 1,379 million b/d on the basis of world GDP at 3.44%. As a downside risk, we highlight the rise in China vehicle sales tax for engines of 1.6 litres displacement or less from 5% to 7.5% which could suppress vehicle sales growth this year.
China oil demand growth is increasingly dependent on private light duty vehicles (PLDV),
at 54% over 2013-16 versus 37% over 2005-08. Our China autos team expects 4.5% growth in vehicle sales, which would equate to PLDV oil demand growth at 324,000 b/d (down from 348,000 b/d last year). However, January sales
were flat year-on-year (yoy) and if annual sales were to stall at zero growth, this would result
in China PLDV oil demand growth at 304,000
In Europe, we expect demand growth to slow, in part as a catch-up to Eurostat figures which showed both imports and gross inland consumption declining as of late last year. In addition, our auto analysts believe vehicle sales growth will slow to 0.5% in 2017 after three years of growth averaging 3.1%. Annual vehicle sales have now reached the replacement rate required to stabilise fleet size which we believe could help European oil demand to stabilise.
This could help to sustain the departure from consistent demand decline of 330,000 b/d yoy during the 2008-13 period, which was matched
by declining vehicle sales averaging 2.3% over
the same period. While the relationship between oil demand growth and vehicle sales is more complex than indicated by a simple plot of the two together, if fleet size remains roughly unchanged then a return to demand decline seems less likely.
In India, the demonetisation of INR500 and 1,000 notes from 8 November 2016 depressed onwards vehicle sales in the three months to January, with a particularly sharp fall of 18.7% yoy in December. The trend in chemical production growth has also continued to slow to 1.4% yoy growth in Q4 of 2016, down from 2.4% in the first three quarters, and 4.2% in 2015.
We expect these factors to drive slower oil demand growth over the course of the year despite high apparent demand growth through January at 656,000 b/d yoy. While India oil demand growth is expected to come more on par with China's over the next decade than it was over the past decade (when it was only 31% of China's growth), the International Energy Agency's projection of 1.6 million b/d cumulative growth from 2016-22 still falls below China at 1.8 million b/d cumulative growth.
Figure 1: World GDP and oil demand growth
Source: IMF, IEA data from Monthly Oil Data Service © OECD/IEA 2017
OPEC members show discipline
We assess compliance against the December
2016 OPEC supply agreement by measuring production only in those countries which committed to reductions (the 10 members apart from Libya, Nigeria and Iran). On this basis, January/February compliance measured 112%, with cuts of 1.4 million b/d achieved versus targeted cuts of 1.2 million b/d. Our assumption is that aggregate compliance for OPEC-10 countries falls back to average 100% over the year, and that OPEC-13 production averages 32.3 million b/d over the year.
For the three non-participating OPEC members, we also saw production falling below our full-year expectations, as Iran failed to produce up to its allocation, militant attacks resumed in Nigeria, and the reopening of a Western pipeline in Libya has been slow to yield results.
Iran remains 32,000 b/d below its output allocation, but the government claims that export capacity has risen to 2.8 million b/d in comparison to total liquids exports of 2.5 million b/d, and that production capacity is around 4 million b/d in comparison to January/February output of 3.7 million b/d. Therefore we expect production will average at the agreed allocation of 3.7 million b/d over the year.
With regard to the potential for a reimposition of US sanctions, we note that Iran has been certified by the International Atomic Energy Agency (IAEA) as being in compliance with the Joint Comprehensive Plan of Action (JCPOA) as recently as 14 February, according to Yukiya Amano, director of the IAEA.
Non-OPEC compliance remains weak
With regard to the companion non-OPEC deal announced on 10 December 2016, we assess compliance in the 11 participating countries at 37%. This is based on the assumption that the baseline level of production is October 2016. We see a 204,000 b/d reduction from the baseline, which is less than the 558,000 b/d targeted reduction. The key shortfall occurred in Russia where the bulk of the targeted cuts reside. A 116,000 b/d reduction was achieved against the targeted 300,000 b/d reduction.
The insignificance of the non-OPEC production cut in January is a function of the fact that production from the countries in question, primarily Russia and Kazakhstan, rose sharply into the October 2016 baseline month. So far, the magnitude of the January cut is almost exactly in line with our business-as-usual assumption for 2017 average production from non-OPEC agreement countries.
Figure 2: Assessing OPEC compliance
* October 2016 for all countries except Angola (September 2016)
** Relative to OPEC secondary source production figure of 3.71 mmb/d
Source: IEA data from Monthly Oil Data Service © OECD/IEA 2017
A surprise from the US
Although we have been well aware of the potential for an upside surprise in US production above our previous model assumption, we had not expected it to come so early in the year. Actual production ran 300,000 b/d above the Drilling Productivity Report-derived tight oil model in December, and this positive divergence has now risen to +500,000 b/d in January. If there is no reconnection between actual figures and the model over the course of the year, our
US supply growth figure could be higher by a similar amount.
It is no secret that one component of the outsized performance is the fact that the oil-directed rig count has grown at a faster rate since October than we have assumed, leading to an average of 498 rigs in the seven tight oil basins in January, versus 427 in our previous model. This outperformance has continued since January, with oil-directed rig additions averaging 11 per week, more than double our assumed run-rate for the year of five per week.
A simple plot of the oil-directed rig count against the three-month lagged West Texas Intermediate (WTI) oil price suggests that we might have foreseen stronger rig additions, and also that the additions likely have further to go if WTI prices simply remain flat at US$52-54 per barrel. In fact, rig activity could well exceed the historical relationship shown since 2014 if the deflation in the cost curve resulting from operational improvements is retained even as activity accelerates, as we believe it will be.
What has been more surprising is the higher revision to historical rig productivity even as rigs are being added back. The expected flattening of rig productivity growth has not yet materialised, although we still expect a slowdown as increasing activity means that more economically marginal acreage is being drilled and less efficient rigs and crews are employed. A retrospective upward revision to rig productivity of roughly 50 b/d stretching back through 2016 has taken place between the November 2016 and February 2017 data sets. If a flattening of productivity is slower to come than we model over the next two years, then upside to production could be even greater.
Figure 3: OECD commercial crude oil inventory
Source: IEA data from Monthly Oil Data Service © OECD/IEA 2017
OPEC members will not have been overly unhappy with the decline in prices towards the bottom of the informal market management band of US$50-60 per barrel described by Iran oil minister Bijian Zanganeh as suitable for most OPEC members. The reason is that the suppression of US supply growth likely remains a secondary goal of policy.
If oil prices remain stable within the prescribed price band, we believe the acceleration in oil-directed rig activity will gradually slow and then plateau by the second half of the year, driven by a balance between rig count and the size of the resource which breaks even at this price level, and reflation in costs as crews are incentivised to come back to work, and older, less technologically capable rigs are pressed back into service.
There is no point in pretending that the persistence of OPEC supply discipline is not the most significant unknown. However, our reading of the situation suggests that, on balance, it is more likely than not that Saudi Arabia leads other members to stay the course not only through H2-17 but also well through 2018. While stronger US production might be cited as a reason why discipline may not be rolled over, we maintain that the consequent deterioration in the market balance, as expressed through OECD commercial crude oil inventory, is consistent with an extension of the Algiers Accord. We note that the OPEC statement in November 2016 which cites an acceleration of "the ongoing drawdown of the stock overhang" and "oil market rebalancing" as the objective of supply policy.
Other key measures will include front month price, of course, but also curve structure which has weakened in both Brent and WTI since the
8 March data release from the US Department
From our point of view the November 2016 statement carries an overly sanguine tone as OECD commercial crude oil inventories have not yet begun any drawdown in earnest, and therefore there is no 'ongoing' drawdown, but rather an 'emerging' drawdown. However, the tone of the statement serves to emphasize how important inventories will be in measuring policy success, and consequently, the conditions under which that policy either must be extended or could be allowed to expire. We believe market conditions will require that policy to be extended through much of 2018.
The impact of an Aramco IPO
As for the significance of the Saudi Arabian Oil Company (Aramco) IPO reportedly planned for 2018, we believe this contributes to the rationale for continued supply discipline, but on its own does not change the calculus in our view. It is not clear to us that the pressures driven by the IPO are any different from the routine challenge of balancing oil revenues against government spending, particularly in a year when government spending is expected to be higher rather than lower.
Moreover, there are voices within the Saudi government who are concerned with selling any portion of the country's 'crown jewels' at what might be viewed as a relatively low valuation, to the extent that the valuation is determined by the current long-term forward oil price. Therefore in the event that Saudi officials viewed it as expedient to dispense with supply discipline prior to the IPO for any reason, we do not see why the IPO could not then be shelved as a consequence.
In short, we believe that the pace of inventory reduction will be slow enough that the maintenance of OPEC supply discipline will be both necessary, as the Saudi national budget rises by 7.9% year-on-year to SAR890bn, and justifiable given broad compliance from both OPEC and non-OPEC producers.
Michael Hsueh and Grant Sporre are research analysists at Deutsche Bank
Already registered? Login to access premium content